Spatial Disorientation in Flight

VISUAL ORIENTATION

Vision is by far the most important sensory modality subserving spatial orientation, especially so in moving vehicles such as aircraft. Without it, flight as we know it would be impossible, whereas this would not necessarily be the case in the absence of the vestibular or other sensory systems that provide orientation information. For the most part, the function of vision in spatial orientation is obvious, so a discussion proportional in size to the importance of that function in orientation will not be presented here. Certain special features of visual orientation deserve mention, however. First, there are actually two separate visual systems, and they have two distinct functions: object recognition and spatial orientation. A knowledge of these systems is extremely important, both to help in understanding visual illusions in flight and to appreciate the difficulties inherent in using flight instruments for spatial orientation. Second, visual and vestibular orientation information are integrated at very basic neural levels. For that reason, spatial disorientation frequently is not amenable to correction by higher-level neural processing.

Anatomy of the Visual System

General

The retina, an evaginated portion of the embryonic brain, consists of an outer layer of pigmented epithelium and an inner layer of neural tissue. Contained within the latter layer are the sensory rod and cone cells, the bipolar and horizontal cells that comprise the intraretinal afferent pathway from the rods and cones, and the multipolar ganglion cells, the axons of which are the fibers of the optic nerve. The cones, which number approximately 7 million in the human eye, have a relatively high threshold to light energy. They are responsible for sharp visual discrimination and color vision. The rods, of which there are over 100 million, are much more sensitive to light than the cones; they provide the ability to see in twilight and at night. In the retinal macula, near the posterior pole of the eye, the cone population achieves its greatest density; within the macula, the fovea centralis--a small pit totally comprised of tightly packed slender cones--provides the sharpest visual acuity and is the anatomic basis for foveal, or central, vision. The remainder of the eye is capable of far less visual acuity and subserves paracentral and peripheral vision.

Having dendritic connections with the rods and cones, the bipolar cells provide axons that synapse with the dendrites or cell bodies of the multipolar ganglion cells, whose axons in turn course parallel to the retinal surface and converge at the optic disc. Emerging from the eye as the optic nerve, they meet their counterparts from the opposite eye in the optic chiasm and then continue in one of the optic tracts, most likely to terminate in a lateral geniculate body, but possibly in a superior colliculus or the pretectal area. Second order neurons from the lateral geniculate body comprise the geniculocalcarine tract, which becomes the optic radiation and terminates in the primary visual cortex, the striate area of the occipital cerebral cortex (Area 17). In the visual cortex, the retinal image is represented as a more or less point-to-point projection from the lateral geniculate body, which receives a similarly topographically structured projection from both retinas. The lateral geniculate body and the primary visual cortex are thus structurally and functionally suited for the recognition and analysis of visual images. The superior colliculi project to the visual association areas (Areas 18 and 19) of the cerebral cortex via the pulvinar, and also eventually to the motor nuclei of the extraocular muscles and muscles of the neck, and appear to provide a pathway for certain gross ocular reflexes of visual origin. Fibers entering the pretectal area are involved in pupillary reflexes. In addition, most anatomic and physiologic evidence indicates that information from the occipital visual association areas, parietal cerebral cortex, and frontal eye movement area (Area 8) is relayed through the paramedian pontine reticular formation to the nuclei of the cranial nerves innervating the extraocular muscles. Via this pathway and perhaps others involving the superior colliculi, saccadic (fast) and pursuit (slow) eye movements are initiated and controlled.

Visual-Vestibular Convergence

Vision in humans and other primates is highly dependent on cerebral cortical structure and function, whereas vestibular orientation primarily involves more primitive anatomic structures. Yet visual and vestibular orientational processes are by no means independent. We know that visually perceived motion information and probably other visual orientational data reach the vestibular nuclei in the brain stem^2,3, but it appears that the integration of visual and vestibular information is to a large extent accomplished in the cerebral cortex of humans.

The geniculostriate projection system is divided both anatomically and functionally into two parts: that incorporating the parvocellular layers of the lateral geniculate body (the "parvo" system) and that incorporating the magnocellular layers (the "magno" system). These systems are largely segregated in the primary visual cortex, undergo further segregation in the visual association cortex, and ultimately terminate in the temporal and parietal lobes, respectively. The parvo system neurons have smaller, more centrally located receptive fields that exhibit high spatial resolution (acuity), and they respond well to color.  They do not, however, respond well to rapid motion or high flicker rates. The magno cells, by comparison, have larger receptive fields and respond better to motion and flicker, but are relatively insensitive to color differences. Magno neurons generally exhibit poorer spatial resolution, although they seem to respond better than parvo neurons at low luminance contrasts. In general, the parvo system is better at detecting small, slowly moving, colored targets located near the center of the visual field, while the magno system is more capable of processing rapidly moving and optically degraded stimuli across larger regions of the visual field.

What is important about these two components of the geniculostriate system is that the parvo system projects ventrally to the inferior temporal areas, which are involved in visual search, pattern recognition, and visual object memory, while the magno system projects dorsally to the posterior parietal and superior temporal areas, which are specialized for motion information processing. The cerebral cortical areas to which the parvo system projects receive virtually no vestibular afferents; the areas to which the magno system projects, on the other hand, receive significant vestibular and other sensory inputs, and are believed to be highly involved with maintaining spatial orientation.

The posterior parietal region projects heavily to cells of the pontine nuclei, which in turn provide the mossy-fiber visual input to the cerebellar cortex. Via the accessory optic and central tegmental tracts, visual information also reaches the inferior olives, which provide climbing fiber input to the cerebellar cortex. The cerebellar cortex, specifically its flocculonodular lobe and vermis, also receives direct mossy-fiber input from its vestibular system. Thus, the cerebellum is another area of very strong visual-vestibular convergence. Furthermore, the cerebellar Purkinje cells have inhibitory connections in the vestibular nuclei and possibly even in the vestibular end-organs; so visual-vestibular interactions mediated by the cerebellum also occur at the level of the brain stem, and maybe even peripherally.

Finally, there is a confluence of visual and vestibular pathways in the paramedian pontine reticular formation. Integration of visual and vestibular information in the cerebellum and brain stem appears to allow visual control of basic equilibratory reflexes of vestibular origin. As might be expected, there also are afferent vestibular influences on visual system nuclei; these influences have been demonstrated in the lateral geniculate body and especially the superior colliculus.

Visual Information Processing

Primary control of the human ability to move and orient oneself in three-dimensional space is mediated by the visual system, as exemplified by the fact that individuals without functioning vestibular systems ("labyrinthine defectives") have virtually no problems with spatial orientation unless they are deprived of vision. The underlying mechanisms of visual orientation- information processing are revealed by receptive field studies, which have been accomplished for the peripheral retina, nuclear relays, and primary visual cortex. Basically, these studies show that there are several types of movement-detecting neurons and that these neurons respond differently to the direction of movement, velocity of movement, size of the stimulus, its orientation in space, and the level of illumination. (For an excellent review of this fascinating topic, see Grǘsser and Grǘsser-Cornehls⁴.)

As evidenced by the division of the primate geniculostriate system into two separate functional entities, however, vision must be considered as two separate processes. Some researchers emphasize the role of the ventral (parvo) system in object recognition (the "what" system) and that of the dorsal (magno) system in spatial orientation (the "where" system); others categorize the difference in terms of form (occipito-temporal) versus motion (occipito-parietal) processing. A recent theory suggests that the dorsal system is primarily involved in processing information in peripersonal (near) space during reaching and other visuomotor activity, whereas the ventral system is principally engaged in visual scanning in extrapersonal (far) visual space.⁵ In the present discussion, we shall refer to the two systems as the "focal" and "ambient" visual systems, respectively, subserving the focal and ambient modes of visual processing. Certain aspects of yet another visual process, the one responsible for generating eye movements, will also be described.

Focal Vision

Liebowitz and Dichgans⁶ have provided a very useful summary of the characteristics of focal vision:

[The focal visual mode] is concerned with object recognition and identification and in general answers the question of "what." Focal vision involves relatively fine detail (high spatial frequencies) and is correspondingly best represented in the central visual fields. Information processed by focal vision is ordinarily well represented in consciousness and is critically related to physical parameters such as stimulus energy and refractive error.

Focal vision uses the central 30 degrees or so of the visual field. While it is not primarily involved with orienting the individual in the environment, it certainly contributes to conscious percepts of orientation, such as those derived from judgments of distance and depth and those obtained from reading flight instruments.

Tredici⁷ categorized the visual cues to distance and depth as monocular or binocular. The monocular cues are (1) size constancy, the size of the retinal image in relation to known and comparative sizes of objects; (2) shape constancy, the shape of the retinal image in relation to the known shape of the object (e.g., the foreshortening of the image of a known circle into an ellipsoid shape means one part of the circle is farther away than another); (3) motion parallax (also called optical flow), the greater displacement of retinal images of nearer objects when an individual is moving linearly in the environment; (4) interposition, the partial obstruction from view of more distant objects by nearer ones; (5) texture or gradient, the apparent loss of detail with greater distance; (6) linear perspective, the convergence of parallel lines at a distance; (7) illumination perspective, which results from the tendency to perceive the light source to be above an object and from the association of more deeply shaded parts of an object with being farther from the light source; and (8) aerial perspective, the perception of objects to be more distant when the image is relatively bluish or hazy.

The binocular cues to depth and distance are (1) stereopsis, the visual appreciation of three-dimensional space that results from the fusion of slightly dissimilar retinal images of an object; (2) vergence, the medial rotation of the eyes and the resulting direction of their gaze along more or less converging lines, depending on whether the viewed object is closer or farther, respectively; and (3) accommodation, or focusing of the image by changing the curvature of the lens of the eye. Of all the cues listed, size and shape constancy and motion parallax appear to be most important for deriving distance information in flying, and they are available at and well beyond the distances at which binocular cues are useful. Stereopsis can provide orientation information at distances up to only about 200 m; it is, however, more important in orientation than vergence and accommodation, which are useless beyond about 6 m.

Ambient Vision

Liebowitz and Dichgans⁶ have also provided a summary of ambient vision:

The ambient visual mode subserves spatial localization and orientation and is in general concerned with the question of "where." Ambient vision is mediated by relatively large stimulus patterns so that it typically involves stimulation of the peripheral visual field and relatively coarse detail (low spatial frequencies). Unlike focal vision, ambient vision is not systematically related to either stimulus energy or optical image quality. Rather, provided the stimulus is visible, orientation responses appear to be elicited on an "all or none" basis The conscious concomitant of ambient stimulation is low or frequently completely absent.

Ambient vision, therefore, is primarily involved with orienting the individual in the environment. Furthermore, this function is largely independent of the function of focal vision. This becomes evident in view of the fact that one can fully occupy central vision with the task of reading while simultaneously obtaining sufficient orientation cues with peripheral vision to walk or ride a bicycle. It is also evidenced by the ability of certain patients with cerebral cortical lesions to maintain visual orientation responses even though their ability to discriminate objects is lost.

While we commonly think of ambient vision as dependent on stimulation of peripheral visual fields, it is more accurate to consider ambient vision as involving large areas of the total visual field, which of course must include the visual periphery. In other words, ambient vision is not so much location- dependent as it is area-dependent. Moreover, ambient vision is stimulated much more effectively by large images or groups of images perceived to be at a distance than by those appearing to be close.

The function of ambient vision in orientation can be thought of as two processes, one providing motion cues and the other providing position cues. Large, coherently moving contrasts detected over a large area of the visual field result in vection, i.e., a visually induced percept of self-motion. If the moving contrasts revolve relative to the subject, he or she perceives rotational self-motion or angular vection (also called circular vection), which can be in the pitch, roll, yaw, or any intermediate plane. If the moving contrasts enlarge and diverge from a distant point, become smaller and converge in the distance, or otherwise indicate linear motion, the percept of self-motion that results is linear vection, which also can be in any direction. Vection can, of course, be veridical or illusory, depending on whether actual or merely apparent motion of the subject is occurring. One can appreciate the importance of ambient vision in orientation by recalling the powerful sensations of self-motion generated by certain scenes in wide-screen motion pictures (e.g., flying through the Grand Canyon in an IMAX theater).

Position cues provided by ambient vision are readily evidenced in the stabilization of posture that vision affords patients with defective vestibular or spinal proprioceptive systems. The essential visual parameter contributing to postural stability appears to be the motion of the retinal image that results from minor deviations from one's desired postural position. Visual effects on posture also can be seen in the phenomenon of height vertigo. As the distance from (height above) a stable visual environment increases, the amount of body sway necessary for the retinal image movement to be above threshold increases. Above a certain height, the ability of this visual mechanism to contribute to postural stability is exceeded, and vision indicates posture to be stable despite large body sways. The conflict between visual orientation information, indicating relative stability, and the vestibular and somatosensory data, indicating large body sways, results in the unsettling experience of vertigo.

One more distinction between focal and ambient visual function should be emphasized. In general, focal vision serves to orient the perceived object relative to the individual, whereas ambient vision serves to orient the individual relative to the perceived environment. When both focal and ambient vision are present, orienting a focally perceived object relative to the ambient visual environment is easy, whether the mechanism employed involves first orienting the object to oneself and then orienting oneself and the object to the environment or involves orienting the object directly to the environment. When only focal vision is available, however, it can be difficult to orient oneself correctly to a focally perceived environmental orientation cue because the natural tendency is to perceive oneself as stable and upright and to perceive the focally viewed object as oriented with respect to the stable and upright egocentric reference frame. This phenomenon can cause a pilot to misjudge his or her approach to a night landing, for example, when only the runway lights and a few other focal visual cues are available for spatial orientation.

Eye Movements

We distinguish between two fundamental types of eye movement: smooth movements, including pursuit, vergence, and those driven by the vestibular system; and saccadic (jerky) movements. Smooth eye movements are controlled at least in part by the posterior parietal cerebral cortex and surrounding areas, as evidenced by functional deficits resulting from damage to these areas. Eye movements of vestibular origin are primarily generated by very basic reflexes involving brain stem mechanisms; and because visual pursuit eye movements are impaired by vestibular and certain cerebellar lesions, the vestibular system appears to be involved in control of smooth eye movements of visual origin. Saccadic eye movements are controlled mainly by the frontal eye fields of the cerebral cortex, which work with the superior colliculus in generating these movements. The frontal eye fields receive their visual input from the cortical visual association areas.

The maintenance of visual orientation in a dynamic motional environment is greatly enhanced by the ability to move the eyes, primarily because the retinal image of the environment can be stabilized by appropriate eye movements. Very powerful and important mechanisms involved in reflexive vestibular stabilization of the retinal image will be discussed in the section dealing with vestibular function. Visual pursuit movements also serve to stabilize the retinal image, as long as the relative motion between the head and the visual environment (or object being observed in it) is less than about 600/sec: targets moving at higher relative velocities necessitate either saccadic eye movements or voluntary head movements for adequate tracking. Saccadic eye movements are used voluntarily or reflexively to acquire a target, i.e., to move it into focal vision, or to catch up to a target that cannot be maintained on the fovea by pursuit movements. Under some circumstances, pursuit and saccadic eye movements alternate in a pattern of reflexive slow tracking and fast-back tracking called optokinetic nystagmus. This type of eye-movement response is typically elicited in the laboratory by surrounding the subject with a rotating striped drum; however, one can exhibit and experience optokinetic nystagmus quite readily in a more natural setting by watching railroad cars go by while waiting at a railroad crossing. Movement of the visual environment sufficient to elicit optokinetic nystagmus provides a stimulus that can either enhance or compete with the vestibular elicitation of eye movements, depending on whether the visually perceived motion is compatible or incompatible, respectively, with the motion sensed by the vestibular system.

Vergence movements, which aid binocular distance and motion perception at very close range, are of relatively minor importance in spatial orientation when compared with the image-stabilizing pursuit and saccadic eye movements. Vergence assumes some degree of importance, however, under conditions where a large visual environment is being simulated in a confined space. Failure to account for vergence effects can result in loss of simulation fidelity: a subject whose eyes are converged to fuse an image representing a large, distant object will perceive that object as small and near. To overcome this problem, visual flight simulators display distant scenes at the outer limit of vergence effects (7-10 meters) or use lenses or mirrors to put the displayed scene at optical infinity.

Even though gross stabilization of the retinal image aids object recognition and spatial orientation by enhancing visual acuity, absolute stability of an image is associated with a marked decrease in visual acuity and form perception⁸. This stability-induced decrement is avoided by continual voluntary and involuntary movements of the eyes, even during fixation of an object. We are unaware of these small eye movements, however, and the visual world appears stable.  

Voluntary scanning and tracking movements of the eyes are associated with the appearance of a stable visual environment, but why this is so is not readily apparent. Early investigators postulated that proprioceptive information from the extraocular muscles provides not only feedback signals for the control of eye movements but also the afferent information needed to correlate eye movements with retinal image movements and arrive at a subjective determination of a stable visual environment. An alternative mechanism for oculomotor control and the subjective appreciation of visual stability is the "corollary discharge" or feed-forward mechanism first proposed by Helmholtz and subsequently by Sperry⁹ and others. Sperry concluded: "Thus, an excitation pattern that normally results in a movement that will cause a displacement of the visual image on the retina may have a corollary discharge into the visual centers to compensate for the retinal displacement. This implies an anticipatory adjustment in the visual centers specific for each movement with regard to its direction and speed." The theoretical aspects of visual perception of movement and stability have been expanded over the years into various models based on "inflow" (afference), "outflow" (efference), and even hybrid sensory mechanisms. The interested reader will enjoy Cohen's concise discussion of these models as they relate to spatial orientation¹⁰.

In developing the important points on visual orientation, we have emphasized the "focal-ambient" dichotomy. As visual science matures further, this simplistic construct will likely be replaced by more complex but valid models of visual processes. Presently we are enthusiastic about a theory in which the dichotomy emphasized is that between the peripersonal (near) and focal extrapersonal (far) visual realms.  This theory argues that the dorsal cortical system and its magno projection pathways are more involved in processing visual information from peripersonal space, while the ventral system and its parvo projections attend to the focal extrapersonal visual environment. The theory also suggests that visual attention is organized to be employed more efficiently in some sectors of three-dimensional visual space than in others (e.g., far vision is biased toward the upper visual field and utilizes local form processing, while near vision is biased toward the lower visual field and is better at global form processing), and that ambient extrapersonal information is largely excluded from attentional mechanisms. Certainly, the current state of knowledge concerning visual orientation is fluid.

VESTIBULAR FUNCTION

The role of vestibular function in spatial orientation is not so overt as that of vision but is extremely important for three major reasons. First, the vestibular system provides the structural and functional substrate for reflexes that serve to stabilize vision when motion of the head and body would otherwise result in blurring of the retinal image. Second, the vestibular system provides orientational information with reference to which both skilled and reflexive motor activities are automatically executed. Third, the vestibular system provides, in the absence of vision, a reasonably accurate percept of motion and position, as long as the pattern of stimulation remains within certain naturally occurring bounds. Because the details of vestibular anatomy and physiology are not usually well known by medical professionals, and because a working knowledge of them is essential to the understanding of spatial disorientation in flight, these details will be presented in the following sections.

Vestibular Anatomy

End-Organs

The vestibular end-organs are smaller than most people realize, measuring just 1.5 cm across. They reside well-protected within some of the densest bone in the body, the petrous portion of the temporal bone. Each temporal bone contains a tortuous excavation known as the bony labyrinth, which is filled with perilymph, a fluid much like cerebrospinal fluid in composition. The bony labyrinth consists of three main parts: the cochlea, the vestibule, and the semicircular canals (Fig. 3). Within each part of the bony labyrinth is a part of the delicate, tubular, membranous labyrinth, which contains endolymph, a fluid characterized by its relatively high concentration of potassium. In the cochlea, the membranous labyrinth is called the cochlear duct or scala media; this organ converts acoustic energy into neural information. In the vestibule lie the two otolith organs, the utricle and the saccule. They translate gravitational and inertial forces into spatial orientation information--specifically, information about angular position (tilt) and linear motion of the head. The semicircular ducts, contained in the semicircular canals, convert inertial torques into information about angular motion of the head. The three semicircular canals and their included semicircular ducts are oriented in three mutually perpendicular planes, thus inspiring the names of the canals: anterior vertical (or superior), posterior vertical (or posterior), and horizontal (or lateral).

The semicircular ducts communicate at both ends with the utricle, and one end of each duct is dilated to form an ampulla. Inside each ampulla lies a crest of neuroepithelium, the crista ampullaris. Atop the crista, occluding the duct, is a gelatinous structure called the cupula (Fig. 4a). The hair cells of which the crista ampullaris is composed project their cilia into the base of the cupula, so that whenever inertial torques of the endolymph ring in the semicircular duct deviate the cupula, the cilia are bent.

Lining the bottom of the utricle is another patch of neuroepithelium, the macula utriculi, whose plane is close to horizontal except for a 20-30° upward slope of its anterior end; and on the medial wall of the saccule in an approximately vertical plane is still another, the macula sacculi (Fig. 4b). The cilia of the hair cells comprising these structures project into overlying otolithic membranes, one above each macula. The otolithic membranes are gelatinous structures containing many tiny calcium carbonate crystals, called otoconia, which are held together by a network of connective tissue. Having almost three times the density of the surrounding endolymph, the otolithic membranes displace endolymph and shift position relative to their respective maculae when subjected to changing gravitoinertial forces. This shifting of the otolithic membrane position results in bending of the cilia of the macular hair cells.

Figure 3. Gross anatomy of the inner ear. The bony semicircular canals and vestibule contain the membranous semicircular ducts and otolith organs, respectively.

The hair cell is the functional unit of the vestibular sensory system. It converts spatial and temporal patterns of mechanical energy applied to the head into neural information. Each hair cell possesses one relatively large kinocilium on one side of the top of the cell and up to 100 smaller stereocilia on the same surface. Hair cells thus exhibit morphologic polarization, that is, they are oriented in a particular direction. The functional correlate of this polarization is that when the cilia of a hair cell are bent in the direction of its kinocilium, the cell undergoes an electrical depolarization, and the frequency of action potentials generated in the vestibular neuron attached to the hair cell increases above a certain resting frequency; the greater the deviation of the cilia, the higher the frequency. Similarly, when its cilia are bent away from the side with the kinocilium, the hair cell undergoes an electrical hyperpolarization, and the frequency of action potentials in the corresponding neuron in the vestibular nerve decreases (Fig. 5).

Figure 4. Vestibular end-organs. a. The ampulla of a semicircular duct, containing the crista ampullaris and cupula. b. A representative otolith organ, with its macula and otolithic membrane.

The same basic process just described occurs in all the hair cells in the three cristae and both maculae; the important differences lie in the physical events that cause the deviation of cilia in the directions in which the various groups of hair cells are oriented. The hair cells of a crista ampullaris respond to the inertial torque of the ring of endolymph contained in the attached semicircular duct as the reacting endolymph exerts pressure on the cupula and deviates it. The hair cells of a macula, on the other hand, respond to the gravitoinertial force acting to displace the overlying otolithic membrane. As indicated in Figure 6a, all of the hair cells in the crista of the horizontal semicircular duct are oriented so that their kinocilia are on the side of the cell facing the utricle. Thus, utriculopetal endolymphatic pressure on the cupula deviates the cilia of these hair cells toward the kinocilia, and all the hair cells in the crista depolarize. The hair cells in the cristae of the vertical semicircular ducts are oriented in the opposite fashion; that is, their kinocilia are all on the side away from the utricle. In the ampullae of the vertical semicircular ducts, therefore, utriculopetal endolymphatic pressure deviates the cilia away from the kinocilia, causing all the hair cells in these cristae to hyperpolarize. In contrast, the hair cells of the maculae are not oriented unidirectionally across the neuroepithelium: the direction of their morphologic polarization depends on where they lie on the macula (Fig. 6b). In both maculae there is a central line of reflection, on opposing sides of which the hair cells assume an opposite orientation. In the utricular macula, the kinocilia of the hair cells are all oriented toward this line of reflection; whereas in the saccular macula, they are oriented away from it. Because the line of reflection on each macula curves at least 90°, the hair cells, having morphologic polarization roughly perpendicular to this line, exhibit virtually all possible orientations on the plane of the macula. Thus, the orthogonality of the planes of the three semicircular ducts enables them efficiently to detect angular motion in any plane; and the perpendicularity of the planes of the maculae plus the omnidirectionality of the orientation of the hair cells in the maculae allow the efficient detection of gravitoinertial forces acting in any direction.

Figure 5. Function of a vestibular hair cell. When mechanical forces deviate the cilia toward the side of the cell with the kinocilium, the hair cell depolarizes and the frequency of action potentials in the associated afferent vestibular neuron increases. When the cilia are deviated in the opposite direction, the hair cell hyperpolarizes and the frequency of action potentials decreases.

Neural Pathways

To help the reader better organize the potentially confusing vestibular neuroanatomy, a somewhat simplified overview of the major neural connections of the vestibular system is presented in Figure 7. The utricular nerve, two saccular nerves, and the three ampullary nerves converge to form the vestibular nerve, a portion of the VIIIth cranial or statoacoustic nerve. Within the vestibular nerve lies the vestibular (or Scarpa's) ganglion, which is composed of the cell bodies of the vestibular neurons. The dendrites of these bipolar neurons invest the hair cells of the cristae and maculae; most of their axons terminate in the four vestibular nuclei in the brain stem--the superior, medial, lateral, and inferior nuclei--but some axons enter the phylogenetically ancient parts of the cerebellum to terminate in the fastigial nuclei and in the cortex of the flocculonodular lobe and other parts of the posterior vermis.

Figure 6. Morphologic polarization in vestibular neuroepithelia. a. All the hair cells in the cristae of the horizontal semicircular ducts are oriented so that their kinocilia are in the direction of the utricle; those hair cells in the cristae of the vertical ducts have their kinocilia directed away from the utricle. b. The maculae of the saccule (above) and utricle (below) also exhibit polarization: the arrows indicate the direction of the kinocilia of the hair cells in the various regions of the maculae. (Adapted from Spoendlin¹¹. )

The vestibular nuclei project via secondary vestibular tracts to motor nuclei of cranial and spinal nerves and to the cerebellum. Because vestibulo-ocular reflexes are a major function of the vestibular system, it is not surprising to find ample projections from the vestibular nuclei to the nuclei of the oculomotor trochlear, and abducens nerves (cranial nerves III, IV, and VI, respectively). The major pathway of these projections is the ascending medial longitudinal fasciculus (MLF). The basic vestibulo-ocular reflex is thus served by sensor and effector cells and an intercalated three-neuron reflex arc from the vestibular nerve to the vestibular nuclei to the nuclei innervating the extraocular muscles. In addition, indirect multisynaptic pathways course from the vestibular nuclei through the paramedian pontine reticular formation to the oculomotor and other nuclei. The principle of ipsilateral facilitation and contralateral inhibition via an interneuron clearly operates in vestibulo-ocular reflexes, and numerous crossed internuclear connections provide evidence of this. The vestibulo-ocular reflexes that the various ascending and crossed pathways support serve to stabilize the retinal image by moving the eyes in the direction opposite that of the motion of the head.

Figure 7. Major connections and projections of the vestibular system.

Via the descending MLF and medial vestibulospinal tract, crossed and uncrossed projections from the vestibular nuclei reach the nuclei of the spinal accessory nerve (cranial nerve XI) and motor nuclei in the cervical cord. These projections form the anatomic substrate for vestibulocollic reflexes, which serve to stabilize the head by appropriate action of the sternocleidomastoid and other neck muscles. A third projection is that from primarily the lateral vestibular nucleus into the ventral gray matter throughout the length of the spinal cord. This important pathway is the uncrossed lateral vestibulospinal tract, which enables the vestibulospinal (postural) reflexes to help stabilize the body with respect to an inertial frame of reference by means of sustained and transient vestibular influences on basic spinal reflexes.

Secondary vestibulocerebellar fibers course from the vestibular nuclei into the ipsilateral and contralateral fastigial nuclei and to the cerebellar cortex of the flocculonodular lobe and elsewhere. Returning from the fastigial and other cerebellar nuclei, crossed and uncrossed fibers of the cerebellobulbar tract terminate in the vestibular nuclei and in the associated reticular formation. There are also efferent fibers from the cerebellum, probably arising in the cerebellar cortex, that terminate not in nuclear structures but on dendritic endings of primary vestibular afferent neurons in the vestibular neuroepithelia. Such fibers are those of the vestibular efferent system, which appears to modulate or control the information arising from the vestibular end-organs. The primary and secondary vestibulocerebellar fibers and those returning from the cerebellum to the vestibular area of the brain stem comprise the juxtarestiform body of the inferior cerebellar peduncle. This structure, along with the vestibular end-organs, nuclei, and projection areas in the cerebellum, collectively constitute the so-called vestibulocerebellar axis, the neural complex responsible for processing primary spatial orientation information and initiating adaptive and protective behavior based on that information.

Several additional projections, more obvious functionally than anatomically, are those to certain autonomic nuclei of the brainstem and to the cerebral cortex. The dorsal motor nucleus of cranial nerve X (vagus) and other autonomic cell groups in the medulla and pons receive secondary vestibular fibers, largely from the medial vestibular nucleus; these fibers mediate vestibulovegetative reflexes, which are manifested in the symptoms of motion sickness (pallor, perspiration, nausea, and vomiting) that can result from excessive or otherwise abnormal vestibular stimulation. Via vestibulothalamic and thalamocortical pathways, vestibular information eventually reaches the primary vestibular projection area of the cerebral cortex, located in the parietal and parieto-temporal cortex. This projection area is provided with vestibular, visual, and somatosensory (proprioceptive) inputs and is evidently associated with spatial orientation processing and with integration of higher-order sensorimotor activity. In addition, vestibular information can be transmitted via long polysynaptic pathways through the brain stem reticular formation and medial thalamus to wide areas of the cerebral cortex; the nonspecific cortical responses to vestibular stimuli that are evoked via this pathway appear to be associated with an arousal or alerting mechanism.

Vestibular Information Processing

As the reader probably deduced while reading the discussion of the anatomy of the vestibular end-organs, angular accelerations are the adequate (that is, physiologic) stimuli for the semicircular ducts, and linear accelerations and gravity are the adequate stimuli for the otolith organs. This statement, illustrated in Figure 8, is the cardinal principle of vestibular mechanics. How the reactive torques and gravitoinertial forces stimulate the hair cells of the cristae and maculae, respectively, and produce changes in the frequency of action potentials in the associated vestibular neurons has already been discussed. The resulting frequency-coded messages are transmitted into the several central vestibular projection areas as raw orientational data to be further processed as necessary for the various functions served by such data. These functions are the vestibular reflexes, voluntary movement, and the perception of orientation.

Figure 8. The cardinal principle of vestibular mechanics: angular accelerations stimulate the semicircular ducts; linear accelerations and gravity stimulate the otolith organs.

Vestibular Reflexes

As stated so well by G. Melvill Jones¹², "...for control of eye movement relative to space the motor outflow can operate on three fairly discrete anatomical platforms, namely: (1) the eye-in-skull platform, driven by the external eye muscles rotating the eyeball relative to the skull; (2) the skull-on-body platform driven by the neck muscles; and (3) the body platform, operated by the complex neuromuscular mechanisms responsible for postural control."

In humans, the retinal image is stabilized mainly by vestibulo-ocular reflexes, primarily those of semicircular-duct origin. A simple demonstration can help one appreciate the contribution of the vestibulo-ocular reflexes to retinal-image stabilization. Holding the extended fingers half a meter or so in front of the face, one can move the fingers slowly from side to side and still see them clearly because of visual (optokinetic) tracking reflexes. As the rate, or correspondingly, the frequency, of movement becomes greater, one eventually reaches a point where the fingers cannot be seen clearly--they are blurred by the movement. This point is at about 60°/sec or 1 to 2 Hz for most people. Now, if the fingers are held still and the head is rotated back and forth at the frequency at which the fingers became blurred when they were moved, the fingers remain perfectly clear. Even at considerably higher frequencies of head movement, the vestibulo-ocular reflexes initiated by the resulting stimulation of the semicircular ducts function to keep the image of the fingers clear. Thus, at lower frequencies of movement of the external world relative to the body or vice versa, the visual system stabilizes the retinal image by means of optokinetic reflexes. As the frequencies of such relative movement become greater, however, the vestibular system, by means of vestibulo-ocular reflexes, assumes progressively more of this function; and at the higher frequencies of relative motion characteristically generated only by motions of the head and body, the vestibular system is responsible for stabilizing the retinal image.

The mechanism by which stimulation of the semicircular ducts results in retinal image stabilization is simple, at least conceptually (Fig. 9). When the head is turned to the right in the horizontal (yaw) plane, the angular acceleration of the head creates a reactive torque in the ring of endolymph in (mainly) the horizontal semicircular duct. The reacting endolymph then exerts pressure on the cupula, deviating the cupula in the right ear in a utriculopetal direction, depolarizing the hair cells of the associated crista ampullaris and increasing the frequency of the action potentials in the corresponding ampullary nerve. In the left ear, the endolymph deviates the cupula in a utriculofugal direction, thereby hyperpolarizing the hair cells and decreasing the frequency of the action potentials generated. As excitatory neural signals are relayed to the contralateral lateral rectus and ipsilateral medial rectus muscles, and inhibitory signals are simultaneously relayed to the antagonists, a conjugate deviation of the eyes results from the described changes in ampullary neural activity. The direction of this conjugate eye deviation is thus the same as that of the angular reaction of the endolymph, and the angular velocity of the eye deviation is proportional to the pressure exerted by the endolymph on the cupula. The resulting eye movement is, therefore, compensatory; that is, it adjusts the angular position of the eye to compensate for changes in angular position of the head and thereby prevents slippage of the retinal image over the retina. Because the amount of angular deviation of the eye is physically limited, rapid movements of the eye in the direction opposite the compensatory motion are employed to return the eye to its initial position or to advance it to a position from which it can sustain a compensatory sweep for a suitable length of time. These rapid eye movements are anticompensatory, and because of their very high angular velocity, motion is not perceived during this phase of the vestibulo-ocular reflex.

Figure 9. Mechanism of action of a horizontal semicircular duct and the resulting reflex eye movement. Angular acceleration to the right increases the frequency of action potentials originating in the right ampullary nerve and decreases those in the left. This pattern of neural signals causes extraocular muscles to rotate the eyes in the direction opposite that of head rotation, thus stabilizing the retinal image with a compensatory eye movement. Angular acceleration to the left has the opposite effect.

 

With the usual rapid, high-frequency rotations of the head, the rotational inertia of the endolymph acts to deviate the cupula as the angular velocity of the head builds, and the angular momentum gained by the endolymph during the brief acceleration acts to drive the cupula back to its resting position when the head decelerates to a stop. The cupula-endolymph system thus functions as an integrating angular accelerometer, that is, it converts angular, acceleration data into a neural signal proportional to the angular velocity of the head. This is true for angular accelerations occurring at frequencies normally encountered in terrestrial activities; when angular accelerations outside the dynamic response range of the cupula-endolymph system are experienced, the system no longer provides accurate angular velocity information. When angular accelerations are relatively sustained or when a cupula is kept in a deviated position by other means, such as caloric testing, the compensatory and anticompensatory phases of the vestibulo-ocular reflex are repeated, resulting in beats of ocular nystagmus (Fig. 10). The compensatory phase of the vestibulo-ocular reflex is then called the slow phase of nystagmus, and the anticompensatory phase is called the fast or quick phase. The direction of the quick phase is used to label the direction of the nystagmus because the direction of the rapid motion of the eye is easier to detect clinically. The vertical semicircular ducts operate in an analogous manner, with the vestibulo-ocular reflexes elicited by their stimulation being appropriate to the plane of the angular acceleration resulting in that stimulation. Thus, a vestibulo-ocular reflex with downward compensatory and upward anti-compensatory phases results from the stimulation of the vertical semicircular ducts by pitch-up (-ay) angular acceleration, and with sufficient stimulation in this plane, up-beating vertical nystagmus results. Angular accelerations in the roll plane result in vestibulo-ocular reflexes with clockwise and counterclockwise compensatory and anticompensatory phases and in rotary nystagmus. Other planes of stimulation are associated with other directions of eye movement such as oblique or horizonto-rotary.

Figure 10. Ocular nystagmus--repeating compensatory and anticompensatory eye movements--resulting from vestibular stimulation. In this case, the stimulation is a yawing angular acceleration to the left, and the anticompensatory, or quick-phase, nystagmic response is also to the left.

As should be expected, there also are vestibulo-ocular reflexes of otolith-organ origin. Initiating these reflexes are the shearing actions that bend the cilia of macular hair cells as inertial forces or gravity cause the otolithic membranes to slide to various positions over their maculae (Fig. 11). Each position that can be assumed by an otolithic membrane relative to its macula evokes a particuiar spatial pattern of frequencies of action potentials in the corresponding utricular or saccular nerve, and that pattern is associated with a particular set of compatible stimulus conditions such as backward tilt of the head or forward linear acceleration. These patterns of action potentials from the various otolith organs are correlated and integrated in the vestibular nuclei and cerebellum with orientational information from the semicircular ducts and other sensory modalities; appropriate orientational percepts and motor activities eventually result. Lateral (ay) linear accelerations can elicit horizontal reflexive eye movements, including nystagmus, presumably as a result of utricular stimulation. Similarly, vertical (az) linear accelerations can elicit vertical eye movements, most likely as a result of stimulation of the saccule; the term elevator reflex is sometimes used to describe this response because it is readily provoked by the vertical linear accelerations associated with riding in an elevator. The utility of these horizontal and vertical vestibulo-ocular reflexes of otolith-organ origin is readily apparent: like the reflexes of semicircular- duct origin, they help stabilize the retinal image. Less obvious is the usefulness of the ocular countertorsion reflex (Fig. 12), which repositions the eyes about their visual (anteroposterior) axes in response to the otolith-organ stimulation resulting from tilting the head laterally in the opposite direction. Presumably, this reflex contributes to retinal image stabilization by providing a response to changing directions of the force of gravity.

Our understanding of the vestibulocollic reflexes has not developed to the same degree as our understanding of the vestibulo-ocular reflexes, although some clinical use has been made of measurements of rotation of the head on the neck in response to vestibular stimulation. Perhaps this situation reflects the fact that vestibulocollic reflexes are not as effective as the vestibulo-ocular reflexes in stabilizing the retinal image, at least not in humans. Such is not the case in other species, however; birds exhibit extremely effective reflex control of head position under conditions of bodily motion--even nystagmic head movements are quite easy to elicit. The high level of development of the vestibulocollic reflexes in birds is certainly either a cause or a consequence of the relative immobility of birds' eyes in their heads. Nonetheless, the ability of a human (or any other vertebrate with a mobile head) to keep the head upright with respect to the direction of applied gravitoinertial force is maintained through tonic vestibular influences on the muscles of the neck.

Vestibulospinal reflexes operate to ensure stability of the body. Transient linear and angular accelerations, such as those experienced in tripping and falling, provoke rapid activation of various groups of extensor and flexor muscles to return the body to the stable position or at least to minimize the ultimate effect of the instability. Everyone has experienced the reflex arm movements that serve to break a fall, and most have observed the more highly developed righting reflexes that cats exhibit when dropped from an upside-down position; these are examples of vestibulospinal reflexes. Less spectacular, but nevertheless extremely important, are the sustained vestibular influences on posture that are exerted through tonic activation of so-called "antigravity" muscles such as hip and knee extensors. These vestibular reflexes, of course, help keep the body upright with respect to the direction of the force of gravity.

Figure 11. Mechanism of action of an otolith organ. A change in direction of the force of gravity (above) or a linear acceleration (below) causes the otolithic membrane to shift its position with respect to its macula, thereby generating a new pattern of action potentials in the utricular or saccular nerve. Shifting of the otolithic membranes can elicit compensatory vestibulo-ocular reflexes and nystagmus, as well as perceptual effects.

Voluntary Movement

It has been shown how the various reflexes of vestibular origin serve to stabilize the body in general and the retinal image in particular. The vestibular system is also important in that it provides data for the proper execution of voluntary movement. To realize just how important such vestibular data are in this context, one must first recognize the fact that skilled voluntary movements are preprogrammed; that is, once initiated, they are executed according to a predetermined pattern and sequence, without the benefit of simultaneous sensory feedback to the higher neural levels from which they originate. The simple act of writing one's signature, for example, involves such rapid changes in speed and direction of movement that conscious sensory feedback and adjustment of motor activity are virtually precluded, at least until the act is nearly completed. Learning an element of a skill thus involves developing a computer-program-like schedule of neural activations that can be called up, so to speak, to effect a particular desired end product of motor activity. Of course, the raw program for a particular voluntary action is not sufficient to permit the execution of that action: information regarding such parameters as intended magnitude and direction of movement must be furnished from the conscious sphere, and data indicating the position and motion of the body platform relative to the surface of the earth--that is, spatial orientation information--must be furnished from the preconscious sphere. The necessity for the additional information can be seen in the signature-writing example cited above: one can write large or small, quickly or slowly, and on a horizontal or vertical surface. Obviously, different patterns of neuromuscular activation, even grossly different muscle groups, are needed to accomplish a basic act under varying spatial and temporal conditions; the necessary adjustments are made automatically, however, without conscious intervention. Vestibular and other sensory data providing spatial orientation information for use in either skilled voluntary or reflexive motor activity are processed into a preconscious orientational percept that provides the informational basis upon which such automatic adjustments are made. Thus, one can decide what the outcome of his or her action is to be and initiate the command to do it, without consciously having to discern the direction of the force of gravity, analyze its potential effects on planned motor activity, select appropriate muscle groups and modes of activation to compensate for gravity, and then activate and deactivate each muscle in proper sequence and with proper timing to accomplish the desired motor activity. The body takes care of the details, using stored programs for elements of skilled motor activity, and the current preconscious orientational percept. This whole process is the major function and responsibility of the vestibulocerebellar axis.

Conscious Percepts

Usually as a result of the same information processing that provides the preconscious orientational percept, one also is provided a conscious orientational percept. This percept can be false, that is, illusory, in which case the individual is said to experience an orientational illusion, or to have spatial disorientation. We can be aware, moreover, that what our bodies tell us about our spatial orientation is not what can be concluded from other information such as flight instrument data. Conscious orientational percepts thus can be either natural or derived, depending on the source of the orientation information and the perceptual process involved; and an individual can experience both natural and derived conscious orientational percepts at the same time. Consequently, pilots who have become disoriented in flight commonly exhibit vacillating control inputs, as they alternate indecisively between responding first to one percept and then to the other.

 

Figure 12. Ocular countertorsion, a vestibulo-ocular reflex of otolith-organ origin. When the head is tilted to the left, the eyes rotate to the right to assume a new angular position about the visual axes, as shown.

Thresholds of Vestibular Perception

Often an orientational illusion occurs because the physical event resulting in a change in bodily orientation is below the threshold of perception. For that reason, the student of disorientation should be aware of the approximate perceptual thresholds associated with the various modes of vestibular stimulation.

The lowest reported threshold for perception of rotation is 0.035°/sec², but this degree of sensitivity is obtained only with virtually continuous angular acceleration and long response latencies (20 to 40 seconds). ¹³ Other observations put the perceptual threshold between roughly 0.1 and 2.0°/sec²; reasonable values are 0.14, 0.5, and 0.5°/sec² for yaw, roll, and pitch motions, respectively.¹⁴ It is common practice, however, to describe the thresholds of the semicircular ducts in terms of the angular acceleration-time product, or angular velocity, which results in just perceptible rotation. This product, known as Mulder's constant, remains fairly constant for stimulus times of about 5 seconds or less. Using the reasonable value of 2°/sec for Mulder's constant, an angular acceleration of 50/sec² applied for half a second would be perceived because the acceleration-time product is above the 2°/sec angular velocity threshold. But a 10°/sec² acceleration applied for a tenth of a second would not be perceived because it would be below the angular velocity threshold; nor would a 0.2°/sec² acceleration applied for 5 seconds be perceived. Inflight experiments have shown that blindfolded pilot subjects are not able to consistently perceive roll rates of 1.0°/sec or less, but can perceive a roll when the velocity is 2.0°/sec or higher. Pitch rate thresholds in flight are also between 1.0 and 2.0°/sec. But when aircraft pitch motions are coupled with compensatory power adjustments to keep the net G force always directed toward the aircraft floor, the pitch threshold is raised well above 2.0°/sec.¹⁵

The perceptual threshold related to otolith-organ function necessarily involves both an angle and a magnitude because the otolith organs respond to linear accelerations and gravitoinertial forces, both of which have direction and intensity. A 1.5° change in direction of applied G force is perceptible under ideal (experimental) conditions. The minimum perceptible intensity of linear acceleration has been reported by various authors to be between 0.001 and 0.03 g, depending on the direction of acceleration and the experimental method used. Values of 0.01 g for a_z and 0.006 g for ax accelerations are appropriate representative thresholds, and a similar value for ay acceleration is probably reasonable. Again, these absolute thresholds apply when the acceleration is either sustained or applied at relatively low frequencies. The threshold for linear accelerations applied for less than about 5 seconds is a more or less constant acceleration-time product, or linear velocity, of about 0.3 to 0.4 m/sec.

Unfortunately for those who would like to calculate exactly what orientational percept results from a particular set of linear and angular accelerations (e.g., those occurring prior to an aircraft mishap), the actual vestibular perceptual thresholds are, as expressed by one philosopher, "constant except when they vary." Probably the most common reason for an orientational perceptual threshold to be raised is inattention to orientational cues because attention is directed to something else. Other reasons might be a low state of mental arousal, fatigue, drug effects, or innate individual variation. Whatever the reason, it appears that individuals can monitor their orientation with considerable sensitivity under some circumstances and with relative insensitivity under others, which inconsistency can itself lead to perceptual errors that result in orientational illusions.

Of paramount importance in the generation of orientational illusions, however, is not the fact that absolute vestibular thresholds exist or that vestibular thresholds are time-varying. Rather, it is the fact that the components of the vestibular system, like any complex mechanical or electrical system, have characteristic frequency responses; and stimulation by patterns of acceleration outside the optimal, or "design," frequency-response ranges of the semicircular ducts and otolith organs causes the vestibular system to make errors. In night, much of the stimulation resulting from the acceleratory environment is indeed outside the design frequency-response ranges of the vestibular end-organs; consequently, orientational illusions occur in flight. Elucidation of this important point is provided in the section entitled "Spatial Disorientation."

Vestibular Suppression and Enhancement

Like all sensory systems, the vestibular system exhibits a decreased response to stimuli that are persistent (adaptation) or repetitious (habituation). Even more important to the aviator is the fact that, with time and practice, one can develop the ability to suppress natural vestibular responses, both perceptual and motor. This ability is termed vestibular suppression.¹⁶ Closely related to the concept of vestibular suppression is that of visual dominance, the ability to obtain and use spatial orientation cues from the visual environment despite the presence of potentially strong vestibular cues. Vestibular suppression seems to be exerted, in fact, through visual dominance because it disappears in the absence of vision.¹⁷ The opposite effect, that of an increase in perceptual and motor responsiveness to vestibular stimulation, is termed vestibular enhancement. Such enhancement can occur (1) when the stimulation is novel, as in an amusement park ride; (2) threatening, as in an aircraft spinning out of control; or (3) whenever spatial orientation is perceived to be especially important. It is tempting to attribute to the efferent vestibular neurons the function of controlling the gain of the vestibular system so as to effect suppression and enhancement, and some evidence exists to support that notion.¹⁸ The actual mechanisms involved appear to be much more complex than would be necessary merely to provide gross changes in the gain of the vestibular end-organs. Precise control of vestibular responses to anticipated stimulation, based on sensory efferent copies of voluntary commands for movement, is probably exercised by the cerebellum via a feed-forward loop involving the vestibular efferent system. Thus, when discrepancies between anticipated and actual stimulation generate a neural error signal, a response is evoked, and vestibular reflexes and heightened perception occur.¹⁹ Vestibular suppression, then, involves the development of accurate estimates of vestibular responses to orientational stimuli repeatedly experienced, and the active countering of anticipated responses by spatially and temporally patterned sensory efferent activity. Vestibular enhancement, on the other hand, results from the lack of available estimates of vestibular responses because of the novelty of the stimulation, or perhaps from a revision in neural processing strategy obligated by the failure of normal negative feed-forward mechanisms to provide adequate orientation information. Such marvelous complexity of vestibular function assures adaptability to a wide variety of motional environments and thereby promotes survival in them.

Thus, the sensory input from the muscle spindles can be biased by descending influences from higher neural centers such as the vestibulocerebellar axis. Thus, the sensory input from the muscle spindles can be biased by descending influences from higher neural centers such as the vestibulocerebellar axis.

Although the muscle spindles are structurally and functionally in parallel with associated muscle groups and respond to changes in their length, the Golgi tendon organs (Fig. 13b) are functionally in series with the muscles and respond to changes in tension. A tendon organ consists of a fusiform bundle of small tendon fascicles with intertwining neural elements, and is located at the musculotendinous junction or wholly within the tendon. Unlike that of the muscle spindle, its innervation is entirely afferent.

The major function of both the muscle spindles and the tendon organs is to provide the sensory basis for myotatic (or muscle stretch) reflexes. These elementary spinal reflexes operate to stabilize a joint by providing, in response to an increase in length of a muscle and concomitant stimulation of its included spindles, monosynaptic excitation and contraction of the stretched agonist (e.g., extensor) muscle and disynaptic inhibition and relaxation of its antagonist (e.g., flexor) muscle through the action of an inhibitory interneuron. In addition, tension developed on associated tendon organs results in disynaptic inhibition of the agonist muscle, thus regulating the amount of contraction generated. The myotatic reflex mechanism is, in fact, the foundation of posture and locomotion. Modification of this and other basic spinal reflexes by organized facilitatory or inhibitory intervention originating at higher neural levels, either through direct action on skeletomotor (alpha) neurons or through stimulation of fusimotor (primarily gamma) neurons to muscle spindles, results in sustained postural equilibrium and other purposive motor behavior. Some researchers have speculated, moreover, that in certain types of spatial disorientation in flight, this organized modification of spinal reflexes is interrupted as cerebral cortical control of motor activity is replaced by lower brainstem and spinal control. Perhaps the "frozen-on-the-controls" type of disorientation-induced deterioration of flying ability is a reflection of primitive reflexes made manifest by disorganization of higher neural functions.

Despite the obvious importance of the muscle spindles and tendon organs in the control of motor activity, there is little evidence to indicate that their responding to orientational stimuli (such as occur when one stands vertically in a I-G environment) results in any corresponding conscious proprioceptive percept.²² Nevertheless, it is known that the dorsal columns and other ascending spinal tracts carry muscle afferent information to medullary and thalamic relay nuclei and thence to the cerebral sensory cortex. Furthermore, extensive projections into the cerebellum, via dorsal and ventral spinocerebellar tracts, ensure that proprioceptive information from the afferent terminations of the muscle spindles and tendon organs is integrated with other orientational information and is relayed to the vestibular nuclei, cerebral cortex, and elsewhere as needed.

Joint Sensation

In contrast to the situation with the so-called "muscle sense of position" just discussed, it has been well established that sensory information from the joints does reach consciousness. In fact, the threshold for perception of joint motion and position can be quite low, as low as 0.5° for the knee joint when moved at greater than 1.0°/sec. The receptors in the joints are of three types, as shown in Figure 13c: (1) lamellated or encapsulated Pacinian corpuscle-like end-organs; (2) spray-type structures, known as Ruffini-like endings when found in joint capsules and Golgi tendon organs when found in ligaments; and (3) free nerve endings. The Pacinian corpuscle-like terminals are rapidly adapting and are sensitive to quick movement of the joint, whereas both of the spray-type endings are slowly adapting and serve to signal slow joint movement and joint position. There is evidence that polysynaptic spinal reflexes can be elicited by stimulation of joint receptors, but their nature and extent are not well understood. Proprioceptive information from the joint receptors projects via the dorsal funiculi eventually to the cerebral sensory cortex and via the spinocerebellar tracts to the anterior lobe of the cerebellum.

One must not infer from this discussion that only muscles, tendons, and joints have proprioceptive sensory receptors. Both lamellated and spray-type receptors, as well as free nerve endings, are found in fascia, aponeuroses, and other connective tissues of the musculoskeletal system, and they presumably provide proprioceptive information to the central nervous system as well.

Cutaneous Exteroceptors

The exteroceptors of the skin include: (1) mechanoreceptors, which respond to touch and pressure; (2) thermoreceptors, which respond to heat and cold; and (3) nociceptors, which respond to noxious mechanical and/or thermal events and give rise to sensations of pain. Of the cutaneous exteroceptors, only the mechanoreceptors contribute significantly to spatial orientation.

A variety of receptors are involved in cutaneous mechanoreception: spray-type Ruffini corpuscles, lamellated Pacinian and Meissner corpuscles, branched and straight lanceolate terminals, Merkel cells, and free nerve endings (Fig. 13d). The response patterns of mechanoreceptors also are numerous: eleven different types of response, varying from high-frequency transient detection through several modes of velocity detection to more or less static displacement detection, have been recognized. Pacinian corpuscles and certain receptors associated with hair follicles are very rapidly adapting and have the highest mechanical frequency responses, responding to sinusoidal skin displacements in the range of 50 to 400 Hz. They are thus well suited to monitor vibration and transient touch stimuli. Ruffini corpuscles are slowly adapting and, therefore, respond primarily to sustained touch and pressure stimuli. Merkel cells appear to have a moderately slowly adapting response, making them suitable for monitoring static skin displacement and velocity. Meissner corpuscles seem to detect primarily velocity of skin deformation. Other receptors provide other types of response, so as to complete the spectrum of mechanical stimuli that can be sensed through the skin. The mechanical threshold for the touch receptors is quite low--less than 0.03 dyne/cm2 on the thumb. (In comparison with the labyrinthine receptors subserving audition, however, this threshold is not so impressive: a O-dB sound pressure level represents 0.0002 dyne/cm2, more than 100 times lower.) Afferent information from the described mechanoreceptors is conveyed to the cerebral cortex mainly by way of the dorsal funiculi and medullary relay nuclei into the medial lemnisci and thalamocortical projections. The dorsal spinocerebellar tract and other tracts to the cerebellum provide the pathways by which cutaneous exteroceptive information reaches the cerebellum and is integrated with proprioceptive information from muscles, tendons, joints, and vestibular end-organs.

Figure 13. Some of the nonvestibular proprioceptive and cutaneous exteroceptive receptors sub serving spatial orientation. a. Muscle spindle, with central afferent (sensory) and more peripheral efferent (fusimotor) innervations. b. Golgi tendon organ. c. Lamellated, spray-type, and free-nerve-ending joint receptors. d. Two of the many types of mechanoreceptors found in the skin: lamellated Pacinian corpuscles and spray-type Ruffini corpuscles.

AUDITORY ORIENTATION

On the surface of the earth, the ability to determine the location of a sound source can play a role in spatial orientation, as evidenced by the fact that a revolving sound source can create a sense of self-rotation and even elicit reflex compensatory and anticompensatory eye movements called audiokinetic nystagmus. Differential filtering of incident sound energy by the external ear, head, and shoulders at different relative locations of the sound source provides the ability to discriminate sound location. Part of this discrimination process involves analysis of interaural differences in arrival time of congruent sounds; but direction-dependent changes in spectral characteristics of incident sound energies allow the listener to localize sounds in elevation and azimuth (and to some extent range), even when the interaural arrival times are not different. In aircraft, binaural sound localization is of little use in spatial orientation because of high ambient noise levels and the absence of audible external sound sources. Pilots do extract some orientation information, however, from the auditory cues provided by the rush of air past the airframe: the sound frequencies and intensities characteristic of various airspeeds and angles of attack are recognized by the experienced pilot, who uses them in conjunction with other orientation information to create a percept of velocity and pitch attitude of the aircraft. As aircraft have become more capable, however, and the pilot has become more insulated from such acoustic stimuli, the usefulness of aircraft-generated auditory orientation cues has diminished.